Pakistan Journal of Science (Vol. 66 No.

1 March, 2014)

ISSN (Print): 0030-9877

ISSN (Online): 2411-0930

IDENTIFICATION OF SHEAR CRACKS IN REINFORCED BEAMS USING FINITE

ELEMENT METHOD (ANSYS)
H.U. Khan, M.N. Rafique, S. Karam, K. Ahmad and A. Bashir
Department of Civil Engineering, University of Engineering and Technology Lahore, Pakistan
Corresponding Author e-mail: hassaan.ullah.khan@hotmail.com

ABSTRACT: Analytical determination of displacements and stresses in reinforced concrete

material was difficult task and engineers had to rely on empirical formulas because concrete consists of
heterogeneous material and creep and shrinkage influenced deformations in it. Due to these
complexities engineers in past had been facing difficulties in coping such problems, but with the
advancement of digital computerization and modern numerical methods for analysis such as finite
element method, these problems can be addressed in a very efficient way. There were two ways to
carry out modeling in ANSYS software, one was smeared approach and the other one was discrete. In
the past, Smeared approach was used to identify the cracks in RC beam using ANSYS but in this work
it was extended using discrete approach of modeling and shear cracks were identified in RC beam and
load deflection curve was simulated which showed good agreement with the experimental results.
Keywords: shear cracks, static analysis, reinforced beam, non linear analysis, simulation.
ultimate uniaxial tensile and compressive strengths are
needed to define a failure surface for the concrete.
Consequently, a criterion for failure of the concrete due
to a multiaxial stress state was calculated by William and
Warnkes (1975) constitutive model for multiaxial
stresses. Bangash (1989) proposed that in a concrete
element, cracking occurs when the principal tensile stress
in any direction lies outside the failure surface. After
cracking, the elastic modulus of the concrete element is
set to zero in the direction parallel to the principal tensile
stress direction. Crushing occurs when all principal
stresses are compressive and lie outside the failure
surface, subsequently, the elastic modulus is set to zero in
all directions and the element effectively disappears.

INTRODUCTION
Concrete structural elements behave differently
under different variety of loading. The identification and
calculation of these responses is very laborious and
requires lot of expense and time. But now a days there are
several techniques available to solve this problem,
amongst those indigenous techniques the widely used one
is finite element method. Finite element method is a
numerical analysis method that divides the structural
element into smaller parts and then simulates static
loading conditions to evaluate the response of concrete
and pre stressed concrete members. The use of this
technique is increasing because of enormous
advancement of engineering and computer knowledge.
This method responds well to non linear analysis as each
component possesses different stress strain behavior. This
behavior is efficiently addressed by software ANSYS
which provides number of elements for modeling of
materials and apply loads to evaluate the response. The
objective of this study was to make a comparison
between experimentally tested RC beam and the one
modeled using ANSYS by incorporating discrete
approach as suggested by Dahmani , etal(2010). The
model beam of Ayman and Banerjee (2007) was taken as
the reference beam for our analysis and shear cracks are
compared with it as obtained from ANSYS.

Finite Element Modeling: Experimental RC beam

specimen was analyzed by using ANSYS which is an
engineering simulation commercially used software
offering a comprehensive suite that spans the entire range
of physics, providing access to virtually any field of
engineering simulation that a design process requires.
The software use its tools to put a virtual product
through a rigorous testing procedure such as testing a
beam under different loading scenarios before it becomes
a physical object. ANSYS can carry out advanced
engineering analyses quickly, safely and practically by
variety of contact algorithms, time based loading features
and nonlinear material models. In this study it used to
carry out discrete modeling of RC beam to analyze it
under static loading conditions.

MATERIALS AND METHODS

Failure Criteria for Concrete: The model developed
using ANSYS is capable of predicting failure for concrete
materials. Both cracking and crushing failure modes are
accounted for. The two input strength parameters i.e.

Reinforced Concrete: For modeling of concrete the

ANSYS used an element named as Solid65 which is non
linear model of brittle material similar to concrete. It was

50

Pakistan Journal of Science (Vol. 66 No. 1 March, 2014)

an eight node solid iso parametric element with three

degrees of freedom at each node.

Fig-1 Reference Beam Dimensions

Table-1 Material Properties for FE model in ANSYS
Material

Density (kg/m3)

Concrete
Reinforcing Steel

2400
7850

Elastic Modulus
(N/m2)
24644.62e6
2e11

Poisons Ratio
0.3
0.2

Fc
(e6 N/m2)
27.54
-

Fy
(e6 N/m2)
280

Table-2 Properties of steel and concrete

Area of main steel
Area of hanger
28 days compressive strength fc

129e-6 m2 (#13)
71e-6 m2 (#10)
27.54e6 N/m2

Steel Reinforcement: for the modeling of steel ANSYS

provided an element named as Link180 There were two
ways to use it one was smeared and the other is discrete,
discrete was considered to be more convergent as it
subtracts the area of steel from total concrete which was
the actual scenario where as in smeared the steel was
embedded in the concrete and behaved as one unit which
was not the actual case.
Experimental Data: Ayman and Banerjee (2007) carried
out their experiment showing the average ultimate failure
load of as-built specimen as 86 kN. The width and height
of the beams tested were 0.160 mm and 0.280m
respectively, the length of the beam was 3.76m with
supports located 0.08m from each end of the beam as
shown in Fig.1. The mild steel flexural reinforcements
used were 2#13 bars, 2#10 hanger bars and shear
reinforcement included #10 U-stirrups. Cover for the
rebar was set to 40mm in all directions.
The experimental failure shown in Fig-2 which is similar
to crack obtained through ANSYS in Fig-12

Fig-2 Shear Crack in Experimentally Tested Beam

The
compressive
uniaxial
stress-strain
relationship for the concrete model was obtained using
implementation of William and Warnkes (1975)
constitutive model for concrete in ANSYS which defined
9 different constants as shown in Table-3

51

Pakistan Journal of Science (Vol. 66 No. 1 March, 2014)

suggested by past researcher Kachlakev and Miller

(2001).

Table-3 Shear Transfer Coefficients

1
2
3
4
5
6
7
8
9

Shear transfer coefficient for an open crack

Shear transfer coefficient for a closed crack
Uniaxial tensile cracking stress
Uniaxial crushing stress (positive)
Biaxial crushing stress (positive)
Ambient hydrostatic pressure
Biaxial crushing stress (positive) under the
ambient hydrostatic stress state
hydrostatic blax
Tensile condition factor

0.3
0.6
1.9e6
-1
0
0
0

Constitutive model for concrete: The following

equations were used to compute the multilinear isotropic
stress-strain curve for the concrete as presented in Fig-3
MacGregor (1992)
f=
E c
(1+(/ o)2)
o= 2fc/Ec
Ec=f/
where:
f = stress at any strain , psi
= strain at stress f
o = strain at the ultimate compressive strength fc
The
multilinear
isotropic
stress-strain
implemented requires the first point of the curve to be
defined by the user. It must satisfy Hookes Law E= /
used by Chote and Barzin (2009)
The following curves are obtained for concrete
and steel in ANSYS V13

0
0.6

Typical shear transfer coefficient range from 0.0

to 1.0, with 0.0 representing a smooth crack (complete
loss of shear transfer) and 1.0 representing a rough crack
(no loss of shear transfer). The shear transfer coefficient
for open and closed cracks was determined using the
work of Kachlakev and Miller (2001) as a basis. The
uniaxial crushing stress in this model was based on the
uniaxial unconfined strength. It was entered as -1 to turn
off the crushing capability of the concrete element as

Fig-3 Stress Strain Curve For Concrete

Fig-4 Stress Strain Curve For Steel

Table-4 showing co-ordinates of beam model in

ANSYS
Table -4 FE beam
Orientation
x1,x2 ,x co-ordinate
y1,y2,y co-ordinate
z1, z2,z co-ordinate

Co-ordinates
0,0.160 m
0,0.280 m
0,1.880 m

By using these co-ordinates the following model

is made as shown in Fig-5

Fig-5 Modeled Beam

52

Pakistan Journal of Science (Vol. 66 No. 1 March, 2014)

After giving material properties for steel and

concrete such as elastic modulus Eb , ultimate uniaxial
compressive strength f c , steel grade, ultimate uniaxial
tensile strength (modulus of rupture) f r , Poissons ratio
for both steel and concrete, shear transfer coefficient, and
compressive uniaxial stress-strain relationship the key
points for the bars are specified to have the shear
stirrups, main and hanger bars as reinforcement are
shown in Fig-6

The stirrup formed using the key point is shown

in Fig-6 and is selected and copied at required spacing
and at required number of times. So finally the beam is
modeled as shown in Fig-7 For the sake of ease due to
symmetry in specimen the right half of the beam is
modeled that requires less time of processing and the side
of the beam in cross section is restrained in the two
perpendicular components only the UY is allowed to
have the degree of freedom as shown in Fig-8, similarly
instead of using metal base plate another concept is
utilized which is fixing the nodes that act as support
providing zero degree of freedom i.e zero displacement.

Fig-8 Support and Side Restraints

RESULTS
After meshing the model, the load is applied in
increments and at different loads, the deflections are
obtained, the stress contour showed the variation of
stresses along the member with deformed shape as shown
in Fig-9 and initial cracking stress contour in presented
Fig-10
ANSYS V13 used to check the response of the
structural member under static load by discrete approach
gave almost the same ultimate load and shear cracks were
identified along with acceptability of ANSYS software to
analyze and predict the cracking pattern which was in
harmony with experimental results. This study also
depicts and explains how to model shear stirrups along
with reinforced bars using half model specimen provided
it is symmetrical. This not just saves the time of iterations
and analysis but also facilitates the user to carry out
modeling quickly. The initial cracking load for both FE
beam and experimental beam was 18.75 KN and the final
cracking load was 86.8KN and 86KN respectively which
was in harmony.

Fig-6 Stirrup and reinforcement

Comparison of FE model and Experimental Results:

The graph is plotted between load and deflection as
obtained through ANSYS and experiment. Results are
shown, before the load of 18.75 KN where initial cracks
appeared, the deflection was same in both experimental
and FE model, but when the initial cracking started the

Fig-7 FE Modeled RC Beam Having Stirrups

53

Pakistan Journal of Science (Vol. 66 No. 1 March, 2014)

curve showed slight variation. It might be because we

have considered a perfect bond between steel and
concrete showing more stiffness and no slippage but in
actual case there was always certain slippage which
might be the reason that FE model was giving lesser
deflection and more stiffness. The initial cracking load
for both FE and experimental beam was 18.75KN which
were in harmony while ultimate load was 86KN for
experimental and 86.8KN for FE model which was also
satisfactory.

At load of about 86.8 KN the member failed,

and crushing of concrete also started at top and near the
support as highlighted in Fig-12

Fig-12 Shear Cracks at Ultimate Load

Fig-13

Fig-9 Deformed Shape

Initial Crack

Conclusion:
1.
ANSYS is time saving and cost efficient tool
that helps in simulation and gives satisfactory results
using discrete approach.
2.
The initial cracking load is same for both FE and
experimental beam.
3.
The difference in the ultimate load might be due
to perfect bonding in steel and concrete in simulated
beam and thus the graph showed high stiffness.

Fig-10 Stress Contour at

The specimen failed at load of 86.8KN and the

deformed shape along with stress distribution is shown in
Fig-11

Recommendation for Future Work: The literature

review and method of modeling used for this study will
not just save time but it also explains the method of
modeling stirrups in FE beam. The bond between
concrete and steel was considered perfect in our analysis
which might be the reason that the FE model was giving
lesser deflection due to more stiffness than actual beam in
which bond was not perfect and certain slippage was
there, therefore, it is recommended to use bond link
element between concrete and steel for future studies for
more accurate results and better calibration.
Acknowledgement: Authors are very grateful to
University of Engineering and technology Lahore,
Pakistan for their moral support.

Fig-11 Deformed Shape and Stress Contour At

Ultimate

54

Pakistan Journal of Science (Vol. 66 No. 1 March, 2014)

Dahmani.L, A. Khennane and S. Kaci Strength of

Materials, Vol. 42, pg 2 (2010)
Kachlakev. D and T. Miller, FE Modeling of Reinforced
Concrete Structures Strengthened with FRP
Lamiates, Final Report SPR 316, Oregon State
University (2001).
MacGregor, J.G. Reinforced Concrete Mechanics and
Design, Prentice-Hall, Inc. Englewood Cliffs,
NJ pg. 25 (1992)
William. K. J. and E. D. Warnke, Constitutive model for
the triaxial behavior of concrete, in Association
for Bridge and Structural Engineering, ISMES,
Vol. 19: pg 1-30,174 (1975).

REFERENCES
Ayman S.M, S. Banerjee Civil and Environmental
Engineering
Department,
University
of
California at Irvine, California 92697,United
States , Elsevier Composites: pg 781793 (2007)
Bangash M. Y. H., Concrete and Concrete Structures,
Numerical Modeling and Applications, Elsevier
Science Publisher Ltd, London, pg 360-370
(1989)
Chote Soranakom and Barzin Mobasher, Flexural Design
Of Fiber Reinforced Concrete, ACI Materials
Journal Vol. 106-M52, pg 2 (2009)

55